Semantic Web Mining:

steelsquareInternet and Web Development

Oct 20, 2013 (4 years and 20 days ago)

87 views

Semantic Web Mining:
Using Association Rules for
Learning an Ontology
Presented By : Amgad Madkour
Agenda

Semantic Web Mining aim

Web Mining overview

Semantic Web overview

Ontology Building

Learning an Ontology using WM
What is the aim ?

To improve, on one hand, the resultsof Web
Mining by exploiting the new semantic
structures in the Web

To make use of Web Mining, on the other
hand, for building up the Semantic Web.
How can both of them help
each other ?

A backbone of semantic web are ontologies
which at present are often hand-crafted.

The challenge is to learn ontologiesand/or
instance of their concepts , in a semi
automatic wayby using web mining
techniques such as association rules.

Conversely , background knowledge –in the
form of ontologies, or in other forms –can be
used to improve the process and results of
web mining.
Web Mining
What is Web Mining ?

Web mining is the application of data mining
techniques to the content, structure, and
usage of Web resources. This can help to
discover global as well as local structure
(“models”or “patterns”) within and between
Web pages

Web mining can profit from given structure on
data (as in database tables), but it can also
be applied to semi-structured or unstructured
data like free-form text
Web Mining Types : Content

Web Content Mining is a form of text mining

The primary Web resource that is being
mined is an individual page.

Web content mining can take advantage of
the semi-structured nature of Web page text.

Web content mining can be used to detect
co-occurrences of terms in texts
Web Mining Types : Structure

Usually operates on the hyperlink structure of Web
pages.

The primary Web resource that is being mined is a
set of pages, ranging from a single Web site to the
Web as a whole.

Exploits the additional information that is (often
implicitly) contained in the structure of hypertext.

Identification of the relative relevance of different
pages that appear equally pertinent when analyzed
with respect to their content in isolation.
Web Mining Types : Usage

The primary Web resource that is being mined is a
record of the requests made by visitors to a Web
site, most often collected in a Web server log.

The actual behavior of those who use these
resources may reveal additional structure than the
one in structure mining.

It is also used for static site improvement by
identifying navigational pattern of the user inside a
Website

It is useful to combine Web usage mining with
content and structure analysis in order to “make
sense”of observed frequent paths and the pages on
these paths
Extracting Semantics from The
Web

The effort behind the Semantic Web is to add
semantic annotation to Web documents in order to
access knowledge instead of unstructured material,
allowing knowledge to be managed in an automatic
way.

Web Mining can help to learn definitions of
structures for knowledge organization (e.g.,
ontologies) and to provide the population of such
knowledge structures.

All approaches discussed are semi-automatic. They
assist the knowledge engineer in extracting the
semantics, but cannot completely replace him/her
Exploiting Semantics for Web
Mining : Web Content Mining

We preprocess the input data (e. g. text) and
apply ontology-based heuristics for feature
selection and feature aggregation.

Based on these representations, we compute
multiple clustering results using k-Means.

The results can be characterized and
explained by the corresponding selection of
concepts in the ontology.
Mining
Semantic Web
What is Semantic Web ?

Huge amount of data is interpretable by
humans only; machine support is limited.

Berners-Lee suggested to enrich the Web by
machine-processableinformationwhich
supports the user in his tasks

To reach this goal the Semantic Web will be
built up in different levels, the one we care
about is ontologies
What is an Ontology ?

An ontologyis a formal explicit description of
conceptsin a domain of discourse ,properties
of each concept describing various features
and instancesof the concept

An ontology together with a set of individual
instancesof classes constitutes a
knowledge base.
Example :
Figure 1. Some classes, instances, and relations among them in the wine
domain. We used black for classesand red for instances. Direct links
represent propertiesand internal links such as instance-of and subclass-of.
Ontology vs. XML

An ontology differs from an XML schema in
that it is a knowledge representation, not a
message format.

The XML specification is not designed to
support reasoningoutside the transaction
context where ontology does.
A Simple Knowledge
Engineering Methodology

There are some fundamental rules in
ontology design:

There is no one correct way to model a
domain

Ontology development is necessarily an
iterative process.

Concepts in the ontology should be close to
objects.
Ontology Building
Step 1: Determine the domain
and scope of the ontology

We suggest starting the development of an
ontology by defining its domainand scope.
That is, answer several basic questions:

What is the domainthat the ontology will cover?

For whatwe are going to use the ontology?

For what types of questions the information in the
ontology should provide answers?

Who will useand maintain the ontology?
Step 2. Consider reusing
existing ontologies

Check if we can refineand extendexisting
sources for our particular domain and task

Reusing existing ontologiesmay be a
requirementif our system needs to interact
with other applications that have already
committed to particular ontologiesor
controlled vocabularies
Step 3. Enumerate important
terms in the ontology

It is useful to write down a list of all terms we
would like either to make statements about or
to explain to a user

It is important to get a comprehensive listof
terms without worryingabout overlap
between concepts they represent, relations
among the terms, or any properties that the
concepts may have, or whether the concepts
are classes or slots..
Step 4. Define the classes and
the class hierarchy

A top-downdevelopment process starts with the
definition of the most general concepts in the
domain and subsequent specialization of the
concepts.

A bottom-updevelopment process starts with the
definition of the most specific classes with
subsequent grouping of these classes into more
general concepts

A combinationdevelopment process: We define
the more salient concepts first and then generalize
and specialize them appropriately.
Example : Concept Hierarchy
Learning an Ontology
Using Web Mining
Step 1 : Extract the Terms from
the Corpus ( Term Extraction )

We obtain the corpusof the domain that we want to
create the ontology for

We start with extracting terms using term extraction
techniques

We specify the linguistic filterusing Part Of Speech
tags

The outputwill be a list of concepts that the
knowledge engineer can include inside his/her
model in order to create the ontology whether in a
automatic or semi automatic way.
Step 2: Apply Association
Rules to the extracted terms

We specify the minimum support level and
minimum confidence level to the terms that
have already been extracted as concepts

The system rechecks the corpus using Part
Of Speech again but now matching the
extracted concepts with the corpus to
determine the association rules between
them

The result then shows the support and
confidence between the resulting concepts
Step 3 : Extract Instances

We can help the knowledge engineer in
automatically or semi-automatically extracting
instances of the concepts that we proposed
so far

The system will not accurately be able to
determine all instances from the corpus there
for the aid of the knowledge engineer is
required in order to obtain a better description
of the ontology
Extra Refining Steps

We can apply pruningto the concepts to
remove any unrelated concepts that could
populate our ontology

As we mentioned the process that we present
is semi automatic and needs the knowledge
engineeringto refine the resulting set of
concepts that the system proposed
Conclusion

A mutual relationship between semantic web
and web mining.

The task of learning ontologiescan be
simplified for the knowledge engineer using
web mining techniques such as association
rules.

We showed a complete run-through on how
to create ontology using the web mining
approach in an automatic/semi-automatic
manner.
Thank You !